Following the broadband money

Posts Tagged ‘infrastructure sharing

How Ofcom encourages network ‘slum landlords’

with 16 comments

Because Ofcom’s regulations do not reflect the underlying technology change from time division multiplexing (TDM) to statistically multiplexed packet switching, those who control access to network-based services can and do behave like “slum landlords”.

This observation is contained in a response to an Ofcom consultation on how its regulations need to change so as to not discourage shared works, shared facilities or revenue sharing and rather support mechanisms for this.

Consulting firm Predictable Network Solutions (PNS) told Ofcom “The market and regulatory structure that has emerged is one that requires the monopolist to charge ‘rent’, and where there is no other measure than the rent that is charged, the result is a race to the bottom where monopolists become ‘slum landlords’ and fitness for purpose is entirely lost.

“Money is charged for access to the resource, not for the resultant outcome, and this attitude persists along the value chain, so that all the accumulated risk (that outcomes will not be achieved) is dumped onto the final customer.”

PNS said its work revealed “an emerging pattern that points to structural flaws in the UK telecoms market. In our view, these flaws constitute a substantive risk that many of the potential benefits to the UK of robust, reliable, ubiquitous and cost-effective telecommunications will not be realised.”

It ascribed this to Ofcom’s traditional view that telecoms is about “deterministic and centrally controlled” circuits rather than packets. “The created retail services are, essentially ‘purpose-for-fitness’, in that ‘you get what you get’; there is no specification of what is delivered, and it is the end users’ problem to find a way of exploiting whatever it is. It is this divergence of expectation from delivery that is creating large scale (and costly) hazards,” it said.

PSN said the current market structure has led to a focus only on the rate of financial return on capital investment. This has been to the detriment of service outcomes.

“Where there is fitness for purpose in the current system (eg TDM), it is fitness for a historical purpose, with no flexibility that would encourage innovation. A further consequence of this ‘rental’ model is there is no market for concurrent services to the end user.”

PNS said customers can switch access providers, but they cannot switch service providers, for example one delivering reliable streaming video, one VoIP and one general web access.

To reduce the risk of owning a network but no “tenants” (think Digital Region), incumbents had to work at “large scale”, and rip off competitors with high prices (think of the fight between BT and Sky over TV content) for access to the network.

This also affects wholesale customers. PNS noted that mobile network operators need to backhaul LTE traffic from small cells. TDM circuits are too costly, so they need to use DSL. “However, the lack of any guarantee of the transport characteristics of such connections (ie mass market DSL) makes their use in a RAN [radio access network] problematic, as the protocols employed were developed to work over TDM.”

PNS called for “an underlying carrier that can offer services that are appropriately isolated from each other and differentiated and guaranteed performance characteristics. A competitive market can then develop in delivering concurrent services over this infrastructure (including, but not limited to, non-discriminatory internet access).”

PNS recommended a regime where service providers offer a set of services with well-defined, quantifiable and verifiable outcomes. These could be offered equally well by competing entities, and other players could built further layers of service on top of them.

Note: Other responders were Ofcom’s Advisory Committee for Wales, Arqiva, the Mobile Operators Association, Virgin Media, Vodafone, and the Wireless Infrastructure Group.


Ex-BT CTO slams government broadband scheme

with 13 comments

Peter Cochrane, former BT head of research and CTO, savaged the government’s broadband policy and implementation in evidence to the House of Lords communications committee which is looking into the UK’s superfast broadband initiative.

Cochrane who left BT in November 2000, said:

In 1979 my PhD was instrumental in BT’s decision to go fibre everywhere.

In 1986 I had got fibre to the home cheaper than copper at 2Mbps.

The island of Jersey is installing a 1000Mbps network everywhere in both directions; it’s cheaper than copper, and they’re doing it because they have a monoculture of banking, and they need to change the economy of the island.

Giving our people 2Mbps is like giving them a Morse key; you might as well not bother.

Twenty to 50 Mbps will not give us entry to cloud computing, on which rests the next phase of industry, commerce and the generation of GDP.

UK broadband is neither super nor fast. For one thing, it’s asymmetrical; communication tends to be two-way, so we can’t make use of video conferencing (with the present system).

When I came in, we were the servants of society…Now…(it’s) I’m here to benefit the company (and) the shareholders. That loss of perspective of a duty to society is really quite damaging.

Fibre to the cabinet is one of the biggest mistakes humanity has made.

What’s this magic about fibre to the home? None. You can put copper Cat5 or LAN cable in yourself. Are you allowed to by BT? I always take the view in everything that I will beg forgiveness later.

Even when I was in the company 85% of UK homes were within one kilometre of a BT fibre that was dark, not being used.

If Ofcom is powerful enough to regulate the radio spectrum, surely they are powerful enough to regulate the waves on fibre.

As we unbundle duct access, there is a case to unbundle fibre and unbundle the wavelengths.

The £560m (£530m actually) that is being talked about is petty cash in this game…(Universal FTTH) will cost about £10bn to £15bn.

I would leverage the £560m (by investing) in small players so that there is a third force. In all successful commercial markets there is a rule of three, perhaps four. What we have right now is a rule of two.

The worst thing I see is start-up companies who get into this space to service people like you and me (in unserved rural areas) who are then observed making a success, and are then wiped out purposely by the incumbent.

If communities take things into their own hands, things can change.

There are two dangers with government investment; either it is spread too thin, or it just impacts one place.

We could have had the same mobile coverage we have today for £2bn if the mobile operators had shared masts. The same is true for fibre networks.

In 1986 I got fibre to the home and by 1990 BT and DuPont had built two factories, one in Ipswich and one in Birmingham… We were rolling out fibre to the home; it was an active programme, but it was stopped…by the Thatcher government and Sir Keith Joseph. They wanted the American cable companies in. The programme was stopped. Working with us were the Japanese and Koreans. They looked on, aghast, as we stopped. They carried on…we went back in time.

Do you take as a general proposition that where there is fibre there should be open access? Yes.

Squeezing the other guy out is not to the benefit of the nation, and not to my mind, a very clever business model.

If you could get (ready) access to fibre near you it would be absolutely transformative.

The cost is in getting the network in; the running cost is relatively low.

The delivery of bandwidth is independent of both the bandwidth and the distance. It’s the reason why your telephone call to North America is worth relatively nothing. It’s the sheer quantity of calls that makes it a viable business.

The analogue trans-Atlantic cable cost about $300m and took about five to seven years to pay back. The first optical fibre cable was filled in about six months and paid back in less than 18 months. Those systems now cost around $350m and pay back in a matter of months, not years.

When I was in BT we had about 7,600 telephone exchanges. We calculated that with fibre, because of its greater reach, we could get it down to under 100, about 60.

The fault level in an optical fibre network is very very low. And you can reduce manning, buildings, power consumption, everything.

A national broadband strategy would be founded on all access for all people at 100Mbps and above with an eye on fair competition, an economic and regulatory framework that encouraged people to help themselves, and encouraged start-up companies to provide the competition that’s necessary.

Written by Br0kenTeleph0n3

2012/03/23 at 00:46

High prices kill BT sub-loop unbundling order test

with 2 comments

BT has pulled the plug on its sub-loop unbundling batch order trial due to lack of demand, which was due to high prices.

In a letter to interested parties BT Openreach’s head of infrastructure products Katherine Roche said she was putting the proof of concept trial “in abeyance” until demand picked up.

Roche said Openreach had not received a “successful order” in the past six months and expected none. She said she would brief the industry soon on the issue.

“The product will not be launched as part of R1800 in Nov(ember) 2011,” she said. R1800 is the next version of Openreach’s 18-month rolling product roadmap.

A BT spokesman said this trial was different to the physical infrastructure access (PIA) trial, the pricing of which has been condemned by communications services providers (CPs). She said Sky, Fujitsu and Callflow were trialling the PIA.

“Our PIA trials have not been suspended. In fact, they are going well,” the spokesman said. “What has been suspended – for now – is a trial of a SLU service where orders could be batched together. This is because no triallists came forward despite being invited to in February.”

They were to test Openreach’s ability to complete five orders on the same day at the same street cabinet as well as CPs’ ability to batch five orders for the same cabinet.

She gave no reason why BT had been unable to attract a “successful order” for the  batch ordering system.  An industry source invited to try the batch order system told Br0kenTeleph0ne, “If it’s too expensive in the first place, it is hardly surprising no-one wants it.”

BT published its price list for PIA earlier this year to howls of protest from the rest of the industry, and warnings that no-one could afford to take it up and still be able to compete with BT.

A consortium led by Fujitsu Telecom, TalkTalk and Virgin Media warned earlier that BT’s prices were four to five times higher than Openreach’s costs, and called for Ofcom, the regulator, to get BT to lower them.

The PIA trial itself was seen by some observers as BT’s ploy to create uncertainty in competitors’ pricing for projects funded by BDUK. BDUK recently announced how it would distribute some £430m to help fund access to high speed broadband in rural areas.

BT has a virtual monopoly in rural areas, and Virgin Media CEO Neil Berkett warned at the weekend that the government risked entrenching BT’s monopoly.

BT struck back in The Times, which reported BT warning “there is a high risk that local councils could waste (half a billion pounds) by backing unsustainable community projects that will not benefit rural communities.”

Sub-loop unbundling costs were also the focus of disputes between Openreach and Thales and Digital Region Ltd, who alleged that BT was overcharging and that Ofcom had erred in setting Ofcom’s prices too high.

Ofcom issued a draft determination in June which found in favour of the plaintiffs, and invited comment on its decision.

Here is the Roche letter:

Subject:SLU Batched Orders Trial – proposal to place in abeyance

To the SLU IWG :

 I am writing to advise that due to insufficient orders, the proof of concept trial for the SLU Batched Order development will be placed in abeyance until such time that sufficient industry demand exists to reconsider. To confirm the product will not be launched as part of R1800 in Nov 2011.

Following indicative prices notified to industry in Jan 2011, interested participants were invited to join the trial in February 2011 with the joint objective to trial Openreach’s ability to complete 5 orders on the same day at the same PCP as well as CPs ability to batch 5 orders for the same PCP. As we have not received a successful order during the past 6 months (and with none forecast), we have been unable to test the concept and our procedures with the Openreach Service and Operation teams. This means we are now unable to satisfy our internal Operational governance procedures in order to launch with R1800 in November.

A briefing to industry will be issued shortly and should you have any enquiries regarding this matter, please contact either myself or your normal SRM

Kind regards

Katherine Roche

Head of Infrastructure Products


(Number removed on request)

Written by Br0kenTeleph0n3

2011/08/31 at 16:02

Posted in News

Tagged with , , ,

The search for fibre’s G spot – continues

with 2 comments

Finding what will trigger large scale investment in fibre networks is like the search for the elusive G spot: there is no shortage of approaches, but most are missing the mark.

This was evident at  Wednesday’s European Competitive Telecommunications Association (ECTA) conference on how to trigger investment in fibre.

The Brussels conference, designed to influence European policy makers, probably left them more confused and despairing of achieving a single market in communications services, and hence missing the aggressive Digital Agenda targets. These are to give all Europeans access to broadband by 2013, and for all to have 30Mbps by 2020, and 50% to be using 100Mbps by then.

The conference showed that each country is responding to the unique circumstances that rule in that country. That response is also heavily influenced by the market position of the incumbent, usually the former state-owned telecom monopoly, and the national regulator’s views of how best to achieve the Digital Agenda targets (whisper this – without relegating the local incumbent either to the scrap heap).

As they like to say in Brussels, the facts on the ground are these: on average incumbents still have 45% of the retail market. They have a virtual monopoly outside the cities. Their main competitors are cable operators, where they exist. Their future competitors, especially in rural areas or “white spots” areas without an adequate broadband service may be mobile operators who use LTE (Long Term Evolution) technology.

Investors think telcos risky investments, at least riskier than other utility-type service providers. To deliver the DA targets telcos must compete in an already tight capital market. Except for video, there is no “killer app” for high speed broadband. Perhaps half the people who now get a 5 to 6Mbps service don’t see the point of faster downloads speed for a premium price, and perhaps 30% of people don’t see the point of the internet at all.

There are huge philosophical differences in the appropriate response to the facts. One of the most interesting new developments is the emergence of the so-called infrastructure fund. These are privately-owned funds that invest “patient money” typically from pension funds, that look for low risk in return for predictable if lower returns.

The pitch to telcos from Henri Piganeau from Cube and Randolph Nijsse of Rabobank’s Communications Infrastructure Fund is simple: you can’t do it with your present balance sheet, so sell us your passive network, sign a long term usage contract with us, and use the cash to introduce innovative services that will draw in millions of subscribers.

Or as Nijsse said bluntly to a stunned Sean Williams, BT’s group director of strategy, policy and portfolio, “How much do you want for your copper network?”

If anyone knows, it should be Williams, a former regulator with Ofcom. But he didn’t say. He has different ideas. In fact he takes issue with almost everything anyone from the continent said.

There’s plenty of money, at least in Britain, Williams says. Besides, Ofcom had priced copper access below cost in the past, and should now let the price rise to drive consumers onto fibre, he says. BT is pricing its fibre-based £2.5bn Infinity service at the same price as copper-based broadband, he says. What’s more, it is spending as much again on copper, largely because FTTC gets higher speeds to consumers faster.

BT regards the linkage between the consumer and the network provider as crucial, but wholesaling the network’s capacity to third parties is key to its business plan, Williams says. Some 45 service providers were already testing the relationship, he says.

Williams’ strongest supporter is cable industry lobbyist Caroline van Weede, MD of Cable Europe. The cable sector is already well-placed to deliver half of the DA targets, and ahead of time, she says. “I think we can leave it to the incumbents to do the rest.”

What should happen to the price of copper access is dividing line. Ofcom economist Peter Culham suggested that pushing it up to drive consumers to fibre is ETNO’s basic position, while driving it down to force incumbents to install fibre is ECTA’s position, he says. He believes it should be held constant to act as an “anchor price” against which to measure alternatives.

The French too are taking a different view. Already 98.5% of the population has access to broadband, despite 20% of them living in 80% of the geographical area. Jerome Coutant of ARCEP says this is because of efforts of the departements (regional authorities). Now the government has said 70% should have fibre access by 2020, and everyone 100Mbps by 2025.

And they are going to encourage, if not force, cooperation between telcos. If one decides to put fibre into an area or a building, it muct tell the others. The others can then decide to buy into the initial risk, or pay a premium for joining later, Coutant says.

As to demand, Coutant says overall consumers want a “bigger story” to go fibre. “People with 20Mbps via ADSL aren’t hungry for more, but take-up in rural areas is in the 50% to 60% range,” he says.

Cut copper prices to boost fibre, regulators told

leave a comment »

The main reason there is so little fibre in Europe is because incumbent network operators are making too much money from their copper networks, and financiers don’t want to upset that cash flow.

If Europe is to achieve the aggressive targets of the Digital Agenda, it must find a way to cut the value of copper and reduce the perceived risk of rebuilding the existing networks with fibre.

This emerged at the European Competitive Telecommunication Association conference on how to kick-start fibre investment in Brussels.

Andreas Weiss, managing director and head of telecommunications, media and technology at investment bank WestLB, said the banks’ preferred borrower was a monopoly with strong, profitable and predictable cash flows. “It’s the top line and then the bottom line that’s important to us,” he told delegates.

Weiss was replying to a question on whether financiers would like to see governments renationalise or create monopoly suppliers of passive broadband infrastructure, a development now underway in Australia.

Karl-Heinz Neumann, general manager and director of WIK Consult, and author of an exhaustive analysis of the effect of wholesale pricing on fibre roll-out, said incumbents were responsible for only 19% of the present investment in fibre, even though they would benefit most by it. “Altnets” (alternative network operators) had spent 73% because they did not have a copper legacy to protect and exploit, he said.

The Digital Agenda calls for 50% of Europeans to have access to 100Mbps broadband by 2020. For this to happen, the wholesale price of copper has to drop from its present average of €8.55 to €6.02 to encourage new build (greenfield) fibre investment, and to €3.42 if the altnet had affordable access to the incumbent’s ducts, poles and other passive infrastructure (a “brownfield” scenario), Neumann said.

If the present access price for copper was firm, fibre network operators would have persuade consumers that the €19.90 access price was worth it, he said.

Neumann called for regulators to give firm and credible timetables to incumbents indicating how and when they would have to cut the wholesale price of copper.

Fibre now reaches just 3.9m (2%) of the total homes in Europe, although it passes 22m, giving a 17% penetration rate. In the UK, BT’s penetration rate is under 3% (144,000 out of 5m homes passed), although it is aiming for 20%.

Why Hunt needs to break the broadband logjam now

with 26 comments

There have been signs, hotly denied by vested interests, that the roll-out of broadband in Britain, may be stalling. Yet an analysis of the numbers suggests that if it isn’t the case, it certainly should be.

Market researcher Point Topic said in May that its broadband infrastructure index had gone backwards, dropping from 55% to 53% over the previous six months. It has cut its forecast for high speed broadband lines in use by 2015 from 8.8m to 6.7m, or just a quarter of the estimated 27m homes and offices in the UK. This is for superfast broadband over existing telephone networks plus new fibre ones, and exclude any that use Virgin Media’s cable network.

This will have come as a blow to culture secretary Jeremy Hunt, who wants Britain to have the “best broadband network in Europe” by 2015. Curiously, an election should take place around then.

Hunt is chasing the European Digital Agenda targets of universal access to a minimum 2Mbps service by 2013, and for a universal 30Mbps service by 2020, with 50% of the population having access to 100Mbps.

He has no chance of matching that unless he acts soon to to break the log jam caused by the twin duopolies of BT and Virgin Media in fixed networks, and MBNL, which supplies T-Mobile, Orange and Three, and Vodacom/O2 on mobiles.

What is the UK’s problem?

Several speakers at the NextGen conference in Essex on 21 June said customers want three things from broadband: availability, reliability, and affordability. In Britain today you can get any two, but not all three at once.

BT, which is leading the charge into next generation broadband with its £2.5bn, mostly fibre to the cabinet (FTTC) Infinity programme, said recently Infinity had passed five million houses. (So, only another 22-23 million to go.)

But that’s quite close to Point Topic’s target. Probably the UK already exceeds it, if you add in Virgin Media’s fibre roll-out, which is mostly to replace or upgrade its cable TV network, and the many, many leased lines to businesses, schools, hospitals, and other public sector premises, which could almost overnight extend further into communities, especially rural ones.

However, BT’s fourth quarter results to March (published in May) showed that the installed base of Infinity users is only 144,000. Kevin McNulty, Openreach’s general manager for next generation access commercial partners, repeated the figure at the NextGen conference.

McNulty said BT was adding between 7,000 and 10,000 Infinity users a week. At that rate, if BT installed no more fibre, it would take more than nine years to occupy the installed services.

Where’s the value?

Perhaps more importantly, the installed base is less than 3% of the available lines. According to McNulty, BT breaks even on its Infinity investment in 12-14 years with a 20% take-up. By comparison, the Norwegian FTTH company Altibox won’t move until 60% of residents have signed up.

Of course, breakeven points also depend on prices. Ofcom, in its wisdom guided by its mandate to look after consumers, has trained everyone to believe that low prices are good. That may have been short-sighted.

McNulty describes 60% of customers as “locked onto” low subscription rates, typically £3.00 to £6.00 a month. Infinity starts at £28 a month with a 40Gb data cap, or uncapped for £38/m. BT also offers an “up to” 20Mbps standard broadband triple play (TV, broadband and phone) for £30. (The Infinity prices include a phone line at £10/m and come with a first three months’ free offer.)

By comparison, Sky offers an up to 20Mbps triple play starting from £7.50/m with the first three months free. And on 10 June Sky announced Sky Go, whereby its subscribers can get free TV on their laptops, smartphones and other mobile devices, and later via Sky’s public network of 4,500 wi-fi hotspots.

According to McNulty, only 20% of BT’s subscribers are on the higher priced packages. It’s “quite a challenge” to get them to upgrade, he says.

The problem with this is that the rest of the market uses BT’s price as the benchmark. As a result, almost all market surveys show that the UK has among the lowest priced, highest penetration and the most used broadband service.

But its average access speeds and take-up of higher speed services are starting to lag competitor nations like Germany and France, and are already well behind the Scandinavian and some former Eastern bloc countries.

Britons clearly feel there is not enough added value in high speed broadband. The government is aware of this. It has hired Martha Lane Fox to get online the 8.7 million (about 13% of the population) who have either not touched a keyboard or who find little to interest them in cyberspace.

But it is also making things worse through things like BDUK’s proposed broadband procurement framework and the “indicative allocations” of BDUK money it will spend through county councils.

Broadband framework will exclude most network operators

Blogger Adrian Wooster, quoting informed sources, says only companies that have sales of more than £40m in the past two years will be able to tender directly under the framework. That excludes even  very capable mid-sized network operators. The department of culture, Olympics, media and sport (DCMS), which owns BDUK, has not responded to requests for confirmation.

However the DCMS officer in charge of BDUK (and from Monday spectrum too), Simon Towler told NextGen delegates the government would announce in July how much money it would give to each of the counties from what remains of the £530m earmarked for rural broadband, about £400m. Split (generously) between 44 counties, excluding the seven that have already got theirs and Cornwall, that’s an average of £12m each.

£12m won’t get broadband to many people even if, as Hunt tells MPs, it is doubled using matching funds, and especially if the UK relies on BT and Virgin Media’s traditional network building methodologies.

There is no shortage of ideas about how to do things differently. In fact the government may be already reeling with confusion at all the suggestions.

As a start, and to show his bona fides, Hunt could do worse than get the Treasury to zero-rate, for the next five years, the business rate tax on lit fibre. And get Ofcom to execute on Parliament’s unanimous vote to extend mobile coverage from 95% to 98% with the upcoming spectrum auction.



In a note received on 27 June, Openreach’s Kevin McNulty makes the following points:

CPs do not want to launch new super fast broadband services in the early days of the roll out, until Openreach has achieved millions of homes passed because of the cost of putting their products and promotions together, a fact that would affect any new local or community entrants. Other major CPs have announced their intention to sell end user services over the Openreach fibre network, thus, numbers are expected to increase dramatically over the next year.

Perhaps more importantly, the installed base is less than 3% of the available lines today, and according to McNulty, BT is targeting a 20% take-up.

McNulty describes the cost comparison as “wrong and mixed up”, saying I do not compare like with like. He says it is right that broadband is price sensitive, but that many of these services are sold as “bundles”, and each contains different elements that make these comparisons confusing. He accepts as “a good point” that there is very little room to charge higher rates for fibre than current broadband services.

BT faces price cuts and refunds on unbundled products

with 4 comments

BT will have to change how much it charges for sub-loop unbundling products and may have to refund money for overcharging, according to a draft determination by the UK’s communications regulator Ofcom.

Ofcom’s heavily redacted draft decision, published on 14 June, relates to a dispute filed by Digital Region Ltd and Thales in February this year after BT’s Openreach subsidiary, which levies the charges, turned down a request to reduce the charges.

The charges have long been a bone of contention between BT and its competitors. The competitors have argued that BT inflated the prices artificially to raise barriers to entry .

Communications minister Ed Vaizey has said he would act on evidence of overcharging by BT.

One other company, whose identity was redacted, joined DLR/Thales in the dispute. Rutland Telecom expressed an interest in the case. It was asked for further information but did not respond, and so was not joined in the case.

DLR is an alternative communications service provider to BT in the south Yorkshire region. It sources the operation of its network services to Thales. It rents some of its physical network, such as the wires between the street cabinet and the customer’s home, from BT.

BT charges “connection fees” of £106.62 and £127.61 respectively where the customer buys both voice and broadband from DLR, or either voice or data from DLR with another provider supplying the other service. BT also charges an annual rent of £93.96 for the connection.

DLR/Thales told Ofcom it estimated the real price of the connection fee should be about £50 in both cases, and that the annual rent should be about £82.35.

DLR/Thales said equivalent prices in New Zealand and the Netherlands were about £50 and £27-28 respectively. However, Ofcom said DLR/Thales had not shown why these prices were relevant in the UK. “We consider that in this case…any international comparison will say very little about BT’s compliance,” it said.

All pricing details have been cut out of Ofcom’s draft determination. This makes it impossible to judge both the financial arguments on both sides and Ofcom’s determination.

This matters because many in the industry believe that BT has “captured” the regulator, a point made by the Guardian newspaper last year.

The parties and any other interested parties have until 5pm on 28 June to submit comments to Ofcom, after which Ofcom will make a final determination.


A BT spokesman said, “We are pleased that DRL/Thales primary claim that the SLU connection charge should be reduced to £50 has not be upheld by the regulator.  Ofcom are proposing that BT make some minor adjustments to our SLU charges although we consider these to be unfounded and we will be responding to Ofcom in due course.”

In a separate dispute, Ofcom turned down a complaint by Opal Telecom (part of TalkTalk) and British Sky Broadcasting  (Sky) that they had overpaid Openreach for some local loop unbundling products.

The regulator acknowledged that the overpayment was due to it miscalculating the fee in 2009, but said a refund would not benefit customers or the level of competition in the market.

The Competition Commission found that Ofcom had “materially erred” by underestimating how fast Openreach could become more efficient; that its assessment of inflation of wage and energy costs was incorrect; and that it had made “certain errors” specifying price caps ancillary services.

Ofcom admitted these mistakes meant Opal and Sky overpaid Openreach.

“However we also note our view that Openreach complied with the regulatory obligations which Ofcom had set. We consider that legal and regulatory certainty in conditions set by Ofcom is important,” Ofcom said.

“We acknowledge that the sums sought by Opal and Sky are not in themselves insignificant, but we also consider that in the context of the overall value of the services provided, any harm to competition that arose during the relevant period was limited, and it is not clear that any retrospective repayment now would remove any alleged competition distortion or economic inefficiency.

“If we were satisfied that a repayment was likely to have a positive effect on consumers or competition, either directly or indirectly, this could alter the balance of relevant factors in favour of ordering a repayment. However, we note in this regard our view that, in the circumstances of these disputes, a repayment would be unlikely to have a significant impact (positive or negative) on competition or consumers.”

The parties have until 29 June to reply.